Heatwave Part II

Happy Sunday. Get ready for a hot one today with temperatures in the 90s throughout the region and upper 90s in spots. My last post contained a discussion of heatwaves and the 6 day heatwave that ended on Thursday. Yesterday, temperatures reached 93 °F, which is almost certain to be the commencement of a heatwave that'll last at least through Monday. I think that Monday will bring in some of the hottest temperatures of the summer so far.

NAM 2-m temperatures for 3 p.m. tomorrow (Tropical Tidbits)

To me, it's felt like heatwaves have become more frequent each year and they've gotten longer. (It's also possible that they are now more overhyped in the media.) Simply put, each year it seems to get more oppressively hot for more sustained periods of time. To see if my speculations are true, I took a look at some historical heatwave data for Bradley International Airport (BDL) in Windsor Locks, CT. The temperature data for BDL date back to 1949.

First, I wanted to see if the length of heatwaves is increasing. Remember, you only need 3 days with the temperature greater than or equal to 90 °F to have a heatwave. I was curious to see if heatwaves are now more frequently 5, 6, or 7 days.

Time-series graph  or heatwave length

The blue regression line of the data for the length of heatwaves at BDL dating back to 1949 has a slope of .000000000121 * 10^-10. With such a small slope for this line, I think we can safely say that heatwaves aren't getting any longer. When first analyzing the data, I figured that this was due to an increase in the number of 3-day heatwaves that we are seeing now. So I decided that I would look to see if we are experiencing more heatwaves each year.

Number of heatwaves each year at BDL

My thoughts that we are seeing more heatwaves each year (including 3-day ones) was somewhat right, since the slope of the blue regression line of the data in the above graph is 0.0148563. This means that each year we are seeing 0.0148563 more heatwaves than the previous year. Curiously, though, the year with the greatest number of heatwaves (7) is actually the first year that data are available for BDL, 1949. I eliminated this datapoint from my analysis to see if there was any change in the above graph.

Number of heatwaves each year at BDL without the 1949 datapoint

Here, we see that the slope of the blue regression line increases to 0.0211603, meaning that there is an average of 0.0211603 more heatwaves each year than the previous year. I found this analysis fun, though I don't think it tells us that we're seeing significantly more heatwaves each year or that they're getting longer. 

Stay cool!